stevehopwoodforex.com
https://www.stevehopwoodforex.com/phpBB3/
Print view

Multi 10:4 Trader EA
https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?t=1619		 Page 73 of 160
Author:  droland [ Sat Apr 06, 2013 2:27 pm ]
Post subject:  Re: Multi 10:4 Trader EA
slipshod wrote:
fx8000 wrote:
slipshod wrote:Another idea - as the NFP comes out at the same time each month, would it be possible to allow the user to specify the hour at which its released (thus allowing for broker GMT+/- differences) and a boolean to specify whether the EA should enter new trades during, say, an hour each side of the report? If the user chooses false then the EA cancels all pendings and doesn't start adding new ones until the NFP is past. Might be useful for those of us who are usually asleep when its released ;)
Many pairs that went into the red yesterday were entered many hours before NFP release. You would have to close all ongoing trades before news release to avoid the catastrophe encountered yesterday.

Unfortunately, there is no easy answer. For me in GMT+ 1 time zone, I would normally stop all trading an hour or two before NFP if there is anything worth trading on NFP day and restart half an hour after news release. Yesterday, letting my live micro account running during NFP release resulted in a big loss. So, lesson learned.
Ok no problem, was just typing ideas off the top of my head as usual :)

Actually my accounts all had losing runs from Wednesday onwards, and NFP was just the icing on the cake of a bad week. The worst was a long on USDCHF at 0.9520, right in time for it to turn and plummet down. Then two longs in succession on GBPCHF that immediately turned and ran for their stops and several AUD based longs (and an EURAUD short) that failed badly.
Reading Nanningbob's thread, he states that he really dislikes trading the CHF pairs for that very reason. There is a lot of unpredictable intervention on the CHF that causes spikes that simply don't fit normal price action.

Similarly he has also been discussing what is currently happening with the Yen. Again, this causes unpredictable price action unlike Non farm Payroll which comes out the first Friday of every month.

That is why it is difficult to write an EA that you simply walk away from and let run on its own. When you discover certain news events that could possibly negatively affect your EA, shut it down or eliminate the pairs that will be most affected.

There is no question we are in a strange market right now. I have to watch my scalper EA's like a hawk but I am still making money. Just not as much.

Dave
Author:  droland [ Sat Apr 06, 2013 2:51 pm ]
Post subject:  Re: Multi 10:4 Trader EA
I agree with the idea that we work on the current EA to get rid of the bugs and improve the entry and exit rules however at some point it might be worthwhile to incorporate some kind of a breakout routine to balance the basket during these unpredictable spikes on the CHF, JPY etc.

London open used to be a favourite for breakout EA's but with the new reality, breakouts can happen any time. Cyprus for example.

I think a breakout routine that watches for unusual breakouts from an ATR that occurs over a relatively short period of time would payoff especially if we were caught on the wrong side of the move of that pair.

Obviously there are several actions we could take:
1. Delete the existing trade and open a trade in the direction of the spike.
2. Leave the existing trade and open a hedge trade of equal value.
3. Open a hedge trade with some factor higher than 1x existing trade.
4. Simple close the existing trade for a loss and leave the rest of the basket trading as normal.

Last week the spikes really hurt. This could minimize the loss and possibly even make a profit.

Dave
Author:  johnf [ Sat Apr 06, 2013 10:33 pm ]
Post subject:  Re: Multi 10:4 Trader EA
Not sure if this helps but this is a news filter indicator that can be configured to show high impact news only. Perhaps the ea can be configured using this?
ffcal.rar
Author:  droland [ Sun Apr 07, 2013 6:13 am ]
Post subject:  Re: Multi 10:4 Trader EA
johnf wrote:Not sure if this helps but this is a news filter indicator that can be configured to show high impact news only. Perhaps the ea can be configured using this?

ffcal.rar
A couple of thoughts on this. This indicator has been around for a long time. Its strength is that it does indicate what currency government will have a major announcement. It doesn't give any indication of what that announcement will be: vis a vis pro or con currency strength, or simply neutral for that matter. Other currencies in our basket, not part of the currency government announcement will not be significantly affected.

If we are going to deal with on going baskets, a spike in our favour on certain currency pairs is positive or beneficial to us. What we really want to do is simply to neutralize or capitalize on a breakout that goes against our established currency pairs (ie pairs that are already in play).

That way we don't have to be out of the market and continue to make profit or minimize losses.

This brings to mind an old adage about analyzing financial statements. There are two ways of increasing profit in any business (and trading is a business):

1. Increases income
2. Decrease expenses

Both affect the bottom line exactly the same.

In our case, we want to decrease our losses (expenses).

Dave
Author:  slipshod [ Sun Apr 07, 2013 6:31 am ]
Post subject:  Re: Multi 10:4 Trader EA
The problem with trading any news event is that you really don't know how far spikes will go & how genuine they are. Sometimes the news is so overwhelmingly positive or negative for the pair that it takes off like a rocket and keeps going. Sometimes its strong news, but big money immediately forces a fake spike in the opposite direction before allowing the real move to occur .. .and sometimes the news is ho-hum and the market reacts with sudden volatile spikes up & down before settling back to where it was.

As this isn't an M5 scalping EA its a bit hard to make any entries on those spikes, as an EA simply can't know whether a given news reaction spike is real, fake then real, or totally fake. The best thing the EA can do is stop trading for that time period. Perhaps it could even have an option to close out all entries on pairs that relate to the news until after the event is past. I'd only suggest that for US unemployment & NFP though, as by & large its the main one that causes large volatile spikes.
Author:  glabbeek [ Sun Apr 07, 2013 7:17 am ]
Post subject:  Re: Multi 10:4 Trader EA
slipshod wrote:The problem with trading any news event is that you really don't know how far spikes will go & how genuine they are. Sometimes the news is so overwhelmingly positive or negative for the pair that it takes off like a rocket and keeps going. Sometimes its strong news, but big money immediately forces a fake spike in the opposite direction before allowing the real move to occur .. .and sometimes the news is ho-hum and the market reacts with sudden volatile spikes up & down before settling back to where it was.
I totally agree with Slipshod. We will have to trade this system for a longer period to see how sensitive it is to news spikes. Last week could have been extreme, or extremely rare and therefore not significant in the long run. We just don't know yet. So far, including last week, the drawdown hasn't been over 7%. So nothing to worry about.

Last week can be looked at in a few different ways.
1 Just a bad week.
2 A problem in the exit. (Too wide, or too tight?)
3 A problem in the entry. (Why didn't the system enter in the opposite direction after being stopped out.)

I would opt for option 1
Author:  sponn [ Sun Apr 07, 2013 7:56 am ]
Post subject:  Re: Multi 10:4 Trader EA
IMO we don't need another s/r, time, etc filters... We just need to find some bugs (I believe that we are almost at the end of this proces).

Why ? Because:
1. ea is very good coded (thx for MrLong and everybody else who have made some improvements)
2. strategy is great (thx to Bob)

Losers are part of the game - that's all. 10.4 strategy is trend following system so you will always get some SL when trends are changing (price is breaking ma240). It is NORMAL with trend following systems. We just need one filter - to avoid taking trades to close to 240 line (already done). No other filters are needed. Last week we just have got major reversals for gbp/eur and back to up trend for all jpy pairs. So price has broken 240 lines causing a buch of losing trades. It will happen again. Each 2-4 months. Period.

Just my 2 cents.
Author:  DING [ Sun Apr 07, 2013 7:59 am ]
Post subject:  Re: Multi 10:4 Trader EA
I prefer Andy option,
>>I'm looking to add check boxes like the buttons next to the pairs, one for long trades and one for short trades, so the trader can decide if he doesn't want to trade a pair then un-tick the box.

I am very appreciate the EA friendly for manual intervention capacity :)

>>IMO we don't need another s/r, time, etc filters... We just need to find some bugs (I believe that we are almost at the end of this proces).

I expect could auto capture the chart to a picture file when each trade have been open and closed,
Author:  pathenry [ Sun Apr 07, 2013 9:12 am ]
Post subject:  Re: Multi 10:4 Trader EA
droland wrote:
pathenry wrote:Can someone please tell me what is wrong with this? I can no longer get the full screen of pairs and it no longer refreshes. Have removed EA, Re booted the computer but nothing changes. Expert reads "cannot load external expert 'stdlib' "
Thanks heaps.
Pat
You have either lost the stdlib.mq4 file in your Empty4 installation or it is corrupted.

Do the following:

1. Close Empty4
2. Go to the experts/libraries subdirectory of your Empty4 installation
3. If you see a copy the stdlib, delete both the .mq4 and .ex4 version of this file
4. copy the attached version into the experts/libraries.
5. restart Empty4 and try you EA again.

Dave
Thankyou Dave, For some reason the EX4 file was missing. Appreciate your help

Pat
Author:  MrLong [ Sun Apr 07, 2013 12:18 pm ]
Post subject:  Re: Multi 10:4 Trader EA
Hi Traders,

Let's see what this week brings, maybe we are at a trend change period like Bob say's in his
manual (page 8).

The bottom line is on average, a trend will change on a currency 2-4 times a year. What that means is 2-6 months of staying with a trend either buying only or selling only.

Looking at a monthly chart, I'm convinced this strategy works..

Andy
All times are UTC Page 73 of 160

Powered by phpBB® Forum Software © phpBB Limited